Designing an enhanced acceptance sampling strategy with the process loss index

Abstract

To reconcile the differences between loose and stringent models and address product specifications sensitive to target value variations, a process loss index Le was introduced. This index assesses process capability by incorporating the concept of the quality loss function. Some researchers have integrated this index into variables sampling plans. In particular, Repetitive Group Sampling Plans (RGSP) offer cost advantages over single sampling plans, but their infinite sampling nature might lead to inefficiencies. To address this, our study introduces a modified RGSP with an adjustable lot disposition mechanism based on the process loss index. We establish an optimisation model to minimise the anticipated number of sample items, considering quality and risk constraints. Performance is compared to conventional methods using metrics such as average sample number, operating characteristic curve, and average run length. Ultimately, a practical demonstration is provided through an example to illustrate and validate the feasibility of the proposed plan. [Submitted: 22 January 2024; Accepted: 17 May 2024]

Access to Document

Other files and links

• Link to publication in Scopus

Fingerprint